Fig 1.
Sampling locations within three vegetation zones in Ghana.
Samples from cattle were collected from multiple sites per vegetation zone, marked with black dots. EVI was extracted from MODIS satellite imagery from the same sites. The capital of Ghana, Accra, is marked with a star.
Fig 2.
Monthly minimum and maximum temperatures, relative humidity and rainfall in the three zones of interest.
a) Minimum (dashed lines) and maximum temperatures (solid line) for the Guinea savannah, deciduous forest, and coastal savannah from 2010 through 2012. b) Percent relative humidity for the Guinea savannah, deciduous forest and coastal savannah from 2010 through 2012. c) Rainfall for the Guinea savannah, deciduous forest, and coastal savannah from 2010 through 2012. These variables are represented by a single weather station in each zone.
Fig 3.
Enhanced Vegetation Index (EVI).
EVI from 2010 through 2012 in the a) Guinea savannah, b) deciduous forest, and c) coastal savannah. EVI was extracted from MODIS satellite imagery at five sites per zone, from which mean (solid line) and +/- 1 standard deviation (dashed lines) statistics were derived.
Fig 4.
Detection of targeted pathogens by multiplex PCR.
M represents the 1Kb plus DNA ladder. BBG+T represents PCR amplicons from simultaneous amplification of B. bigemina and Theileria species. AM+BBG represents PCR amplicons from simultaneous amplification of A. marginale and B. bigemina. AM+T PCR represents amplicons from simultaneous amplification of A. marginale and Theileria species. AM+BBG+T represents PCR amplicons from simultaneous amplification of A. marginale, B. bigemina and Theileria species.
Table 1.
Prevalence of targeted pathogens by vegetative zone.
Fig 5.
The percentage of animals acquiring more than one pathogen within a twelve-week period.
The x-axis represents the percentage of animals infected with more than one pathogen. The y-axis represents the time of sampling in weeks.
Table 2.
Prevalence predicted by the best multiple linear regression models for each pathogen as determined by best subsets variable selection and minimum Akaike Information Criterion corrected (AICc) for small sample size.
Table 3.
Model averaging based on corrected Akaike Information Criterion (AICc) weight (Wt).